So I'm working on a document processing AI and I already have a character recognition model which performs decently well. Now the problem is, how do I feed each character to the model in order to make predictions. Sliding window is one technique to segment arbitrary sequences of the image but it might not work in many cases as the words may get cut off. Can anyone suggest a robust way to detect text regions in an image? I've been researching about MSER but they claim to work well on grayscale images and not RGB. I know that we can convert RGB to grayscale but that is not performant enough.